Production AI • LLM Cost Optimization
Start here
Start with this
Removing the LLM From the Hot Path: Building SThree’s Skills Taxonomy Pipeline
13 min read
A 30-experiment programme that replaced a per-skill LLM validator with a zero-LLM retrieval-and-filter cascade and a 7 kB classifier head — 180× cheaper, ~100× faster, and within ~2pp of the model it replaced.
18 May 2026
Enter article